Introduction

As a refugee resettlement caseworker, my interest in this UNHCR dataset is rooted in my professional experience. I worked with clients from Afghanistan and Ukraine who arrived in the United States through temporary immigration pathways, prompting many to seek asylum. This sparked my curiosity about whether there are discernible patterns regarding the denial of asylum applications among individuals from specific countries of origin.

Expectations:

• Anticipated that Ukraine might have higher asylum acceptance rates than Afghanistan; stemmed from belief that racism and Islamophobia still persist in the United States
• Expected applicants from Venezuela, Haiti, and Cuba to have higher asylum rejection rates compared to Afghanistan and Ukraine
• Expectation based on the nature of their asylum claims,less media attention paid to their situations, & historically less government support in creating permanent residency pathways for migrants from these nations

Importance: It is crucial to comprehend this issue within the context of global migration management and to uncover any significant disparities between groups granted asylum and those who are not. Ensuring the fairness and equity of the asylum case system is important, especially considering that individuals applying for asylum are already in vulnerable situations. This makes it even more important that the system remains unbiased and does not discriminate against individuals from specific countries.

Furthermore, it is important to understand this issue from both a legal aid and social service perspective. Analysis of this issue might indicate:

• need for more targeted outreach to individuals from specific countries of origin to assist with their asylum applications

• signal need for reforms within the United States asylum system

• signal need for increased lobbying efforts in Congress to establish legal pathways to permanent residency for citizens from these countries

Method/Data:

The data is sourced from the United Nations High Commissioner on Refugees and encompasses asylum case application data from 2015 to 2023. However, for the purpose of this analysis, I chose to filter the data to:

• include only the years 2021 to 2023
• only include the variables ‘Year’, ‘Country of Origin’, ‘Country of Asylum’, ‘Rejected Decisions’, and ‘Total Decisions’.

For the purposes of this analysis I created a new variable, entitled “Rejected.percent.overall”, to represent the percentage of cases for asylum applicants from each country of origin that were rejected.

Rejected.percent.overall = ‘Rejected Decisions’ + ‘Otherwise Closed Decisions’ / sum of ‘Total Decisions’ * 100. It represents the percentage of asylum case rejections for each ‘Country of Asylum’ and ‘Country of Origin’ pair.

Filtering by Country: I chose to only include applicants from:

• Afghanistan
• Cuba
• Haiti
• Ukraine
• Venezuela

Afghanistan and Ukraine were selected due to the significant displacement resulting from ongoing conflicts in those regions. Additionally, I included Haiti, Venezuela, and Cuba because these nationalities have constituted the bulk of asylum applications in the United States over the past two years.

I utilized the summarize function following grouping by Country.of.Asylum, Country.of.origin, and Year to calculate the total number of asylum applications and rejections by the country where the application was filed, categorized by the country of origin of asylum-seekers.

Results

In terms of the results of the regression, the base ‘Country of Origin’ which each other ‘Country of Origin’ is being compared to is Afghanistan. The results indicate that for every other ‘Country of Origin’, they have statistically significant higher asylum case rejection rates than those from Afghanistan. This can be observed in the visual below.

The visual depicts the asylum rejection rates by ‘Country of Origin’ using the parameters of the regression. The black line running through each box indicates the median asylum rejection percentage, with the bottom of the box indicating the 1st quartile, or 25% of the data, and the top of the box indicating the 3rd quartile of the data, or 75%.

This visual indicates that:

• Asylum seekers from Cuba are the most likely to have their asylum cases rejected, with a median asylum rejection rate of~77%
• Asylum applicants from Haiti have a median asylum rejection rate of 71%
• Asylum applicants from Venezuela have a median asylum rejection rate of ~ 60%
• Asylum applicants from Ukrainian have a median asylum rejection rate of ~58%
• Asylum applicants from Afghanistan have a median asylum rejection rate of ~40%

These findings are interesting to me because I was expecting Ukrainian applicants to have the lowest asylum case rejection rates. Additionally, I was not expecting both Cuban and Haitian asylees to have such high asylum application rejection rates.

Findings

The regression has shown a statistically significant relationship between the country of origin of asylum applicants and the percentage of asylum cases which are rejected. Specifically, Afghan asylum seekers have significantly lower asylum rejection rates than applicants from other countries of origin. This is exemplified by the p-value for every country of origin aside from Afghanistan being below 0.001. P-values by country were:

• Cuba: < 2e-16
• Haiti: < 2e-16
• Ukraine: 6.01e-06
• Venezuela: 3.91e-06

This indicates that the chances of observing these changes in asylum rejection rates, if there was no relationship between the country of origin being Afghanistan and a different country of origin, would be less than 0.001.

This is interesting to me and makes me curious what factors make applications for Afghan asylees more likely to be approved. Reasons for this occurring could be:

• legal grounds for their asylum applications
• greater ability to access adequate legal services
• applicants eligibility for deportation under a “safe third-party agreement”